Nuclear Magnetic Resonance Spectroscopy: A Multifaceted Toolbox to Probe Structure, Dynamics, Interactions, and Real-Time In Situ Release Kinetics in Peptide-Liposome Formulations.

Liposomal formulations represent attractive biocompatible and tunable drug delivery systems for peptide drugs. Among the tools to analyze their physicochemical properties, nuclear magnetic resonance (NMR) spectroscopy, despite being an obligatory technique to characterize molecular structure and dynamics in chemistry as well as in structural biology, yet appears to be rather sparsely used to study drug-liposome formulations. In this work, we exploited several facets of liquid-state NMR spectroscopy to characterize liposomal delivery systems for the apelin-derived K14P peptide and K14P modified by Nα-fatty acylation. Various liposome compositions and preparation modes were analyzed. Using NMR, in combination with cryo-electron microscopy and dynamic light scattering, we determined structural, dynamic, and self-association properties of these peptides in solution and probed their interactions with liposomes. Using 31P and 1H NMR, we characterized membrane fluidity and thermotropic phase transitions in empty and loaded liposomes. Based on diffusion and 1H NMR experiments, we localized and quantified peptides with respect to the interior/exterior of liposomes and changes over time and upon thermal treatments. Finally, we assessed the release kinetics of several solutes and compared various formulations. Taken together, this work shows that NMR has the potential to assist the design of peptide/liposome systems and more generally drug delivery systems.

Protein residue linking in a single spectrum for magic-angle spinning NMR assignment.

Here we introduce a new pulse sequence for resonance assignment that halves the number of data sets required for sequential linking by directly correlating sequential amide resonances in a single diagonal-free spectrum. The method is demonstrated with both microcrystalline and sedimented deuterated proteins spinning at 60 and 111 kHz, and a fully protonated microcrystalline protein spinning at 111 kHz, with as little as 0.5 mg protein sample. We find that amide signals have a low chance of ambiguous linkage, which is further improved by linking in both forward and backward directions. The spectra obtained are amenable to automated resonance assignment using general-purpose software such as UNIO-MATCH.